On the convergence of recursive algorithms with constant gain matrix used in adaptive systems
Roberto Moura Sales · International Journal of Control · 1985
In this paper we study the convergence of a recursive algorithm with constant gain matrix commonly used in adaptive systems. This work was suggested by the consensus of simulation examples carried out by the author or published in the literature, which show that this algorithm converges, despite the absence of analytical results. We show initially, by an analytical example, that the algorithm does not converge in general. Then we propose an alteration in the algorithm so that the resulting algorithm corresponds to the original one implemented in a digital computer. We show that this new algorithm converges, so explaining the simulation results.